CD70–CD27 Amplifies Atypical B Cell Polarization Without Defining the Suppressive State

Question: How do CD27 and CD70 signaling regulate the polarization and suppressive function of atypical memory B cells in chronic inflammatory and tumor microenvironments?

Answer

Across the retrieved papers, CD70–CD27 emerges as a required but non-sufficient amplifier of atypical memory B cell polarization, and not as the switch that defines their suppressive identity. The only causal evidence comes from acute myeloid leukemia, where atypical B cells are expanded, phenocopy the atypical B cells of chronic inflammation, and require Toll-like receptor, interferon and CD70–CD27 signaling for polarization — a process further amplified by CAR T cell–AML contact, with the resulting suppressor output delivered through PD-1 and TGF-β rather than IL-10 [2]. In chronic inflammatory and infection settings, CD27 instead behaves largely as a stage marker: atypical B cell differentiation proceeds through progressive CD24 and CD27 loss, terminating in cells lacking CD24, CD27 and CD21 that carry the antigen-presentation and effector programs [3], and circulating tissue-like/atypical populations in HIV are themselves defined as CD21−CD27− [1]. Taken together, these papers place CD70 responsiveness at an early-to-intermediate, CD27-retaining stage of the atypical pathway, while the terminally differentiated, most tissue-engaged cells have shed the receptor — a decoupling that no retrieved study has directly tested and that makes "CD27 signaling regulates suppressive function" currently a claim about polarization rather than about effector suppression.

CD70–CD27 is one of three convergent inputs required for atypical B cell polarization, and CAR T cells feed it

In AML bone marrow, atypical memory B cells were expanded across single-cell and flow cytometry cohorts, resembled atypical B cells of chronic inflammatory conditions, and were polarized within the leukemic microenvironment, with polarization amplified by CAR T cell–AML interactions. Mechanistically, that polarization was dependent on Toll-like receptor, interferon and CD70–CD27 signaling [2]. The same study shows the loop is self-reinforcing: the polarized B cells suppressed T cell and CAR T cell responses, were enriched in non-responding patients with AML and myeloma, and could be circumvented by CD180-directed CAR T cells that engage the shared Batyp/blast antigen [2]. The claim supported here is that CD70–CD27 is embedded in a cytokine/TLR network rather than operating as a standalone instructive signal, and that the tumor is an active participant in generating its own suppressive B cell niche.

The suppressive module installed by that polarization is PD-1/TGF-β-based, not IL-10-based

Polarized Batyp suppressed T cell and CAR T cell responses via PD-1 and TGF-β signaling [2]. Other routes to suppressive B cells converge on different effectors: a CD40-targeted IL-21 fusokine combined with TLR9 activation generated human B cells expressing IL-10 and granzyme B in >95% of input cells, which suppressed activated T cell proliferation and modulated inflammatory myeloid responses [7], and aged lung accumulates PD-L1+CTLA-4+IgM+IgD− regulatory B-like cells that directly impair NK cytotoxicity and CD4 T cell responses [8]. The claim is that "suppressive atypical B cell" is not a single effector state, and that the module attributed to CD70–CD27-dependent polarization is checkpoint- and cytokine-based.

CD27 is progressively lost along the atypical B cell differentiation axis, so CD27 marks the input-responsive stage rather than the terminal suppressor

High-dimensional flow cytometry, single-cell transcriptomics and lineage analysis in human malaria resolved atypical B cell differentiation into a trajectory marked by progressive CD24 and CD27 loss, culminating in late atypical B cells lacking CD24, CD27 and CD21 with high CD11c, T-bet and NKG7, partly overlapping double-negative 2 cells; early and intermediate stages retained greater plasma-cell and cytokine potential, whereas later stages acquired antigen-presentation signatures [3]. Tissue-like memory cells in HIV are CD21−CD27−, express multiple inhibitory receptors, and split into T-bet-low and T-bet-high subsets differing in differentiation stage, isotype and mutational burden [1]. The pathogenic compartment targeted therapeutically in rheumatoid arthritis is likewise defined as CD27−IgD−Zeb2+CD11c+T-bet+ DN2 [5]. The claim: the population that is phenotypically most suppressive and most tissue-engaged is the one that has lost CD27, which confines any CD70–CD27 requirement to an earlier window.

Costimulation dependence of this compartment is therapeutically demonstrable, but the axis implicated in chronic inflammation is B7–CD28

In a prospective open-label rheumatoid arthritis trial, abatacept-induced low disease activity correlated with coordinated depletion of circulating peripheral helper T cells, late-activated naive cells, and CD27−IgD−Zeb2+CD11c+T-bet+ DN2 cells, with DN2 cells upregulating CXCR4 in a manner that reversed on drug withdrawal and was interpreted as egress from synovial infiltrates [5]. Conversely, DN2-like IgD−CD27−CD11c+CD21− cells can be generated in vitro from healthy donor B cells using a cocktail targeting BCR, TLR7, TLR9 and CD40 — with no T cell or CD70 source [6]. The claim is that costimulatory dependence of this compartment is real in patients, but the retrieved chronic-inflammatory evidence points to B7–CD28 and CD40 rather than CD70–CD27 as the actionable costimulus.

Stromal, damage-derived and interferon cues can install suppressive atypical B cell states without documented CD70–CD27 involvement

Aged CD140a adventitial lung fibroblasts drive CXCL13-dependent reprogramming of B cells into a senescent regulatory B-like state marked by Il10, Ctla4, Cd274 and Egr1 upregulation, which promotes pulmonary colonization by disabling NK and CD4 responses [8]. In obesity, CD11c+T-bet+ B cells expand in spleen, adipose tissue and liver, and T-bet-deficient B cells show enhanced germinal center, Tfh and Tph responses with higher antibody affinity — consistent with these cells constraining germinal center output during chronic inflammation [9]. Apoptotic bodies from patients with Sjögren's disease expand atypical memory B cells and drive autoantibody production, an effect attenuated by blocking BCR and TLR7-9 signaling [4]. In refractory SLE, durable remission after CD19 CAR T therapy depended on truncating the extrafollicular trajectory including CD11c+T-bet+ DN2 cells, and relapse was preceded by elevated CXCL13 and interferons with expanded PD1hi T peripheral helper cells [10]. The claim: multiple microenvironmental inputs — stromal chemokines, apoptotic cargo, interferon, metabolic inflammation — converge on the same atypical state, so CD70–CD27 is best read as one entry route into a shared program.

Conflicts and gaps

The central conflict is directional and cellular: the AML study reports dependence on "CD70-CD27 signaling" without specifying whether CD70 is displayed by CAR T cells, blasts, or the B cells themselves, or whether CD27 signals into the B cell versus CD70 reverse-signaling into the partner [2]. A second conflict is requirement versus dispensability: in vivo polarization in AML required CD70–CD27 [2], yet DN2-like cells are generated in vitro by BCR/TLR/CD40 stimulation alone [6], and the trajectory that culminates in CD27-negative atypical B cells is reconstructed from infection and healthy-donor data that did not perturb the CD70–CD27 axis at all [3]. A third is marker versus mediator: one literature uses CD27 loss to define terminal differentiation [3], while another uses CD27 engagement to define the polarization requirement [2]; whether late CD27-negative atypical B cells remain CD70-responsive is untested. Fourth, the effector module assigned to CD70–CD27-polarized cells (PD-1/TGF-β) is not compared head-to-head with IL-10/granzyme B or PD-L1/CTLA-4 programs within the same population [2] [7] [8]. Finally, every chronic-inflammatory claim is correlative or inferred from costimulation blockade that does not target CD70–CD27 [5], and no retrieved paper tests CD70 blockade or CD27 agonism on atypical B cell polarization in RA, SLE, Sjögren's, HIV or obesity.

What would change the answer

Cell-type-resolved conditional genetics in the AML model — deleting Cd70 separately in T cells, blasts, myeloid cells and B cells, and Cd27 in B cells — would settle directionality and necessity [2]. A human polarization assay that sorts CD27-retaining versus CD27-negative atypical B cell stages along the malaria-defined trajectory and titrates CD70-Fc or blocking antibody against TLR/interferon stimulation, reading out PD-1 and TGF-β, would directly test whether CD27 loss confers refractoriness [3] [2]. Proximity labeling or spatial transcriptomics at Batyp–CAR T interfaces in marrow, and in RA or SLE synovium, would establish whether CD70–CD27 engagement occurs in situ. Clinical pharmacodynamics in patients receiving CD70-directed or CD27-directed agents — paired single-cell RNA/ATAC with BCR sequencing plus soluble CD27 to distinguish transcriptional downregulation from shedding — would convert the current correlative chronic-inflammation data into a test of causality [5] [1], and direct comparison of the PD-1/TGF-β, IL-10/granzyme B and PD-L1/CTLA-4 modules within sorted atypical B cells from the same tissue would resolve whether these are alternative or sequential states.

References

[1] Muir, L., Suchanek, O., Thomas, P., Griffith, S. A., Touizer, E., Rees-Spear, C., Yoshida, M., Pinder, C. L., Nikolic, M. Z., Doores, K. J., Van Gils, M., Gupta, R. K., Clatworthy, M. R., McCoy, L. E.. "Multifaceted B-cell response to transient HIV viraemia in elite controllers". bioRxiv. 2024-12-28T00:00:00.000Z. URL: https://doi.org/10.1101/2024.12.28.630596

[2] Gottschlich, A., Kavaka, V., Janert, T. A., Igl, V., Carlini, E., Rohrbacher, L., den Hartog, W., Harwood, J., Trefny, M. P., Sogukcelik, S., Penovici, P., Ransmayr, B., Rausch, C., Brauneck, F., Frenz-Wiessner, S., Goek, I., Haenel, G., Zucchinetti, C., Oner, A., Zwicker, P. S., Gruenmeier, R., Delahais, S., Schnell, P. M., Nandi, S., Thomas, M., Brabenec, R., Briukhovetska, D., Michaelides, S., Mueller, R. D., Ballheimer, L. Y., Tsiverioti, C. A., Hoffmann, G. V., Stock, S., Piseddu, I., Mueller, P. J., Kruppa, M., Fiedler, W., Rade, M., Kreuz, M., Kehl, N., Bhagwat, A. S., Thiel, J. T., Ste. "A CAR T cell-amplified suppressive B cell circuit in AML". bioRxiv. 2026-10-19T00:00:00.000Z. URL: https://doi.org/10.64898/2026.10.12.758644

[3] Courey-Ghaouzi, A.-D., Kleberg, L., Mousavian, Z., Lautenbach, M. J., Pirronello, M., Phad, G. E., Forsell, M. N., Färnert, A., Sundling, C.. "CD24 and CD27 resolve ontogeny and gradual differentiation of CD11c+ atypical memory B cells after malaria infection". bioRxiv. 2026-07-05T00:00:00.000Z. URL: https://doi.org/10.1101/2025.09.15.676235

[4] Lee, A. Y. S., Shi, B., Gummadi, S., Fonseka, P., Tu, T.-K. M., Le, Q. T., Ang, C.-S., Nguyen, T. K., Lin, M. W., Phan, T. K., Reed, J. H.. "Apoptotic bodies from patients with Sjogren's disease drive atypical memory B cell and macrophage activation and autoantibody production". bioRxiv. 2026-07-20T00:00:00.000Z. URL: https://doi.org/10.64898/2026.07.15.738605

[5] Shwetar, J. J., Amarnani, A., Rigby, W., Skopelia-Gardner, S., Ruggles, K. V., Silverman, G. J.. "Costimulatory blockade depletes T peripheral helper, late-activated naive, and DN2 B cells in rheumatoid arthritis". medRxiv. 2026-03-16T00:00:00.000Z. URL: https://doi.org/10.64898/2026.03.14.26348386

[6] Aguiar, V. R. C., Franco, M. E., Abdel-Aziz, N., Fernandez-Salinas, D., Chinas, M., Colantuoni, M., Xiao, Q., Hackert, N., Liao, Y., Cervantes-Diaz, R., Todd, M., Wauford, B., Wactor, A., Prahalad, V., Laza-Briviesca, R., Darbousset, R., Wang, Q., Jenks, S., Cashman, K. S., Zumaquero, E., Zhu, Z., Case, J., Cejas, P., Gomez, M., Ainsworth, H., Marion, M., Benamar, M., Lee, P., Henderson, L., Chang, M., Wei, K., Long, H., Langefeld, C. D., Gewurz, B. E., Sanz, I., Sparks, J. A., Meidan, E., Nigrovic, P. A., Gutierrez-Arcelus, M.. "A multi-omics resource of B cell activation reveals genetic mechanisms for immune-mediated diseases". medRxiv. 2025-06-14T00:00:00.000Z. URL: https://doi.org/10.1101/2025.05.22.25328104

[7] Galipeau, J., Pennati, A., Yuan, S., Das, R., Hedican, C., Yeware, A.. "A CD40-Targeted IL-21 Fusokine Enables Rapid Generation of Human IL-10⁺Granzyme B⁺ Regulatory B Cells". bioRxiv. 2026-01-13T00:00:00.000Z. URL: https://doi.org/10.64898/2026.01.12.699061

[8] Chen, C., Zuo, W., Huang, C., He, J., Chen, H., Shi, J., Ren, G.. "Aging-associated regulatory B-like cells establish a lung-selective immunosuppressive niche to promote pulmonary metastasis". bioRxiv. 2026-08-18T00:00:00.000Z. URL: https://doi.org/10.64898/2026.08.10.743922

[9] Vanz, C., Enslow, B., Collins, E., Dominguez-Lowry, M., Liendo, N., Dudley, E. A., Leadbetter, E. A.. "CD11c+ Tbet+ B cells constrain obesity- and vaccination-induced germinal center B cells and T helper cells". bioRxiv. 2025-10-02T00:00:00.000Z. URL: https://doi.org/10.1101/2025.09.01.673552

[10] Noethling, D.-M., Anoshkin, K., Gavin, P. G., Rothe, T., Bucci, L., Iwata, F., Garantziotis, P., Ferrari, N., Clarke, S. L. N., Hagen, M., Wirsching, A., Bachl, J., Tur, C., Boeltz, S., Kretschmann, S., Aigner, M., Voelkl, S., Munoz, L., Mueller, F., Mackensen, A., Eckstein, M., Raimondo, M. G., Bozec, A., Schett, G., Grieshaber-Bouyer, R.. "A Feed-Forward Loop Between Extrafollicular B Cell Differentiation and the Inflammatory Milieu Governs Remission and Relapse in Systemic Lupus Erythematosus". medRxiv. 2026-07-13T00:00:00.000Z. URL: https://doi.org/10.64898/2026.07.03.26356448